The key to Taobao store operation: select and measure the store’s models!

　　For sellers, choosing and testing models is a very important link. If you choose the model well, the traffic and sales after it is launched. How to choose and test models on Taobao? Before we start promoting, we need data to analyze how this baby is and whether it is competitive in the market. Through the direct train, we can see data on click-through rate, collection purchase rate, conversion rate, etc. to analyze it, do a good job in the internal strength of the baby (main picture, details, basic sales and evaluation, etc.), and do a good job in related sales of the baby to improve conversion rate.

　　The test model is comprehensive, accurate and direct, and it is nothing more than using a direct train to test the model. Let’s briefly talk about the test process:

　　1. The daily limit can be selected from 200-300, and the style should not be too many to keep at 3-5 models. Standard delivery, the PC side can be turned off but not closed, so the bid will be set. Both the targeted and off-site are closed, and the time discount is 100%. For example, it can be not placed at 0-8 am, and the region will not be placed in remote places.

　　2. After creating the plan, try to select more accurate long-tail words for keywords, try not to add large words first, do not open targeted and intelligent matching, you can add four creative titles, try to select consistent creative titles, try to cover the main keywords promoted as much as possible, and try to drive the weight and start the carousel of creativity.

　　3. Keyword bidding depends on your category. If you want non-standard products, you can bid below the average industry price. If you want to get a higher premium, try to get a standard product, and at least you have a certain amount of traffic. I won’t mention this here. According to the daily limit, you must first ensure the display, look at the data for 1-3 days. If there is no display, increase the price and obtain the display.

　　4. There must be a certain amount of traffic as a reference basis for the test model, and at least 100-200 traffic can be determined. The more traffic, the more accurate the data is. The shopkeeper with sufficient funds can have a higher daily limit.

　　5. Finally, based on the data in the express train, refer to the collection rate, purchase rate, and conversion rate of the baby, and test the words, pictures, and people that are suitable for the baby. The collection and purchase rate mentioned here needs to be calculated by yourself.

　　But how to operate it? Let’s briefly talk about it below.

　　First, choose words

　　I have seen many great masters say that they have spoken of, and there are many channels, and I have also said that in previous posts. Then how do we usually choose the word? Here are some commonly used ones:

　　1. Business consultant selects words (some words for natural flow conversion)

　　2. Some more representative words for babies

　　3. Category words on the homepage

　　4. System recommendation

　　5. Pull down box

　　6. Direct train backend-traffic analysis

　　7. Competitor's conversion words

　　8. Baby attributes, drivers themselves form words

　　9. Auxiliary software (such as Magic Mirror, Shop Chacha, etc.)

　　Because we are trying to test the patterns, we must not choose too many words in terms of word selection. Try to choose some more accurate words with search index, but if it is too long-tail, such as very accurate long-tail words, the conversion rate of such words will be higher, which may affect the later analysis of data. In the later daily promotion, we do not only recommend accurate words, but also need some words with relatively large traffic to promote.

　　Second, time discount

　　Before promotion, we must understand the daily traffic trend of the category and choose the appropriate delivery time. For example, if the baby has no foundation and the weight is relatively low, you can choose a time period with the average traffic situation. If you want to get the data test faster, you can choose to place it in the peak traffic period of one day.

　　Let me talk about the time discount. In this regard, we mainly set it based on the transformation of the baby. There is a conversion rate trend in the entire industry in the business consultant. If the store is not partly the same type of products, it can be set based on this. There is also a line chart on the homepage of our express train, which can also be used as a reference point. The reference data, like test data, must be based on a period of time and must also have a certain amount of traffic to avoid accidents and inaccuracies.

　　Of course, if the store data is small, you can choose an industry template to use it first, and adjust it according to the data when the later data is sufficient.

　　Third, wireless terminal

　　Now we all know that in most categories, wireless terminals have accounted for a large proportion of the entire network traffic, which can also be seen during last year's Double Eleven. The data on all aspects of the wireless side must be excellent with the PC side, so when it is deployed, we can bid on the wireless side, and even without putting it on the PC side to obtain the most traffic. When choosing keywords on the wireless side, you can select words with mobile phone tags and words recommended in the system recommendations. The words recommended in the drop-down box are all composed of core words and attribute words. It can also be added to the title for title optimization. There is also the wireless terminal details part, which must be made separately and should not be used for PC terminal details.

　　Fourth, region

　　When you place the region, you can use it according to your own category. Which region has a large demand for our baby is worth promoting, and what region has a small demand for our baby is not put into it. In addition, for example, some seasonal babies, similar down jackets, are definitely larger in the northern region than in the south, so when considering regional issues, you should refer to seasonality.

　　Of course, you can adjust it based on the data in the later stage. If the daily limit is relatively small, you can choose several regions. How to choose? In the direct train backend - in traffic analysis, you can refer to the keywords that are recommended. Of course, there are also places to refer to in business consultants.

　　Fifth, keyword post-optimization

　　As mentioned earlier, when we tested in the early stage, we tried not to select too many words, which could be kept within 30. It is easier to adjust through data feedback when there are not many keywords to ensure that the analysis of the data is correct. After a certain period of time, we have to look at the data of each word. Good words can increase bids, bad words can be bargained or even deleted, and then add some words for testing and screening. It won't take long.

　　In the later stage, targeted regional and time can be made through the express train. The specific region and time of placement need to be considered by yourself to see what products you are making. If you don’t know how to operate it at the beginning, you can choose a few models for testing first. After all, driving is a technical job, and you can only gain experience after trying it. Finally, let’s summarize, the test model is actually very simple and not that complicated. This is basically the case for the test model. We must consider the life cycle of the baby. The alternation of each season may affect the baby. The main purpose of testing is to confirm that the baby is suitable for the main promotion. It is suitable for us to focus on promotion. To lay the groundwork for the peak season, whether a store can continue to develop is nothing more than the maintenance of one hit product after another.